Re: [PATCH v2 11/12] perf/x86: make HT bug workaround conditioned on HT enabled

[Stephane Eranian]

On Wed, Oct 22, 2014 at 3:27 PM, Jiri Olsa <jolsa@xxxxxxxxxx> wrote:
> On Thu, Oct 09, 2014 at 06:34:45PM +0200, Stephane Eranian wrote:
>
> SNIP
>
>> + */
>> +static __init int fixup_ht_bug(void)
>> +{
>> +     int cpu = smp_processor_id();
>> +     int w, c;
>> +     /*
>> +      * problem not present on this CPU model, nothing to do
>> +      */
>> +     if (!(x86_pmu.flags & PMU_FL_EXCL_ENABLED))
>> +             return 0;
>> +
>> +     w = cpumask_weight(topology_thread_cpumask(cpu));
>> +     if (w > 1) {
>> +             pr_info("CPU erratum BJ122, BV98, HSD29 worked around\n");
>> +             return 0;
>> +     }
>> +
>> +     watchdog_nmi_disable_all();
>> +
>> +     x86_pmu.flags &= ~(PMU_FL_EXCL_CNTRS | PMU_FL_EXCL_ENABLED);
>> +
>> +     x86_pmu.commit_scheduling = NULL;
>> +     x86_pmu.start_scheduling = NULL;
>> +     x86_pmu.stop_scheduling = NULL;
>> +
>> +     watchdog_nmi_enable_all();
>
> if you build with CONFIG_LOCKUP_DETECTOR=n it wont link:
>
>   LD      init/built-in.o
> arch/x86/built-in.o: In function `fixup_ht_bug':
> /root/linux/arch/x86/kernel/cpu/perf_event_intel.c:3267: undefined reference to `watchdog_nmi_disable_all'
> /root/linux/arch/x86/kernel/cpu/perf_event_intel.c:3275: undefined reference to `watchdog_nmi_enable_all'
>
> I think you need watchdog_nmi_(enable|disable)_all stubs out of kernel/watchdog.c
>
Good catch. In fact the stub cannot be in watchdog.c because it may
not even get compiled. Needs to be
in watchdog.h. I fixed that now.
Thanks
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/